fix(jobs): explain missing positional arguments (WBC-266)

[ClayMav]

Missing job-run arguments only reported a count, leaving users to guess what to supply. Create, logs, and metrics now name the required argument and show usage, an example, and a help command. Blank and extra arguments receive the same guidance.

Fixes WBC-266.

Verification

• make fmt, make test, make build, and git diff --check passed.
• Added command execution tests for missing, blank, and extra inputs across all three commands. Existing successful command tests pass.
• Ran the built CLI against a local OpenAPI fixture: missing logs/create inputs and extra metrics inputs exited 1 with the expected guidance. job-runs logs run-123 --dry-run and job-runs logs --help exited 0. Only the local spec endpoint received a request.
• Independent local agent reviewed commit 52045e4 with no findings and reran the argument-error and successful-creation tests; both passed.
• GitHub Test and Build, semantic-title-check, and Salty Hambot review passed on the same commit. Salty reported no findings. No live Cloud requests were made.

Authored with GPT-6 in the Codex harness.

[ClayMav]

@salty-hambot review

[salty-hambot]

Reviewed by Salty Hambot 🤖🧂 — rubric mode

Verdict: ✅ pass

Dimension	Verdict	Notes
correctness	✅ pass	Arg validation refactor is sound; UseLine() correctly emits [flags], tests confirm behavior.
security	✅ pass	No auth, injection, or secret-handling changes in the touched paths.
privacy	✅ pass	No PII or data handling introduced; only CLI arg-validation error text.
reliability	✅ pass	Validator handles missing, blank, and extra-arg cases deterministically.
scalability	✅ pass	N/A — pure CLI argument parsing, no scaling surface.
observability	✅ pass	Error messages are more actionable now (usage + example + help hint).
clarity/maintainability	✅ pass	Consolidates three duplicated inline checks into one reusable helper — a net readability win.
test quality	✅ pass	Table-driven TestJobsArgumentErrors covers all three commands x missing/blank/extra cases with usage-line assertions.

Clean, well-tested refactor that swaps ad-hoc arg checks for a shared validator with actionable usage hints. Reviewer B's 'missing [flags]' scare is a false alarm — Cobra's UseLine() adds [flags] automatically when the command has flags, which all three do. Ship it.

_{0 finding(s) posted · 1 filtered as false positives.
💰 Review cost: $2.6400 · 794.1k in / 6.1k out tokens · ⏱️ 1m49.1s
💬 To request a re-review, comment @salty-hambot review}